Erythropoietin (EPO) and erythropoiesis-stimulating agents (ESAs) are used to treat the anemia of chronic renal disease in greater than 90% of all in-center hemodialysis patients at a cost of approximately 2 billion dollars per year. Despite protocols for anemia management in the end stage renal disease (ESRD) population, a large proportion of patients do not respond predictably to typical doses of EPO. Several recent randomized controlled trials looking to increase hemoglobin (Hb) in patients with the anemia of renal disease have uncovered many questions about the treatment of anemia with ESAs not previously addressed in new drug applications or in subsequent research. A call has been made for the establishment of the optimal Hb target, dosing algorithm, and monitoring approach for patients with anemia from chronic renal disease. We suggest that our poor understanding of the mechanisms leading to EPO resistance prevents the ability to objectively predict and design dosing algorithms for anemia management. We propose that this problem is best addressed with a semi-empirical method which uses surrogate markers of EPO responsiveness and computer-directed algorithms to achieve a specific Hb target range and to exaggerated oscillations in Hb levels. The broad-based long term goals of our research are to address a critical barrier to the progress with personalization of anemia management by developing more objective, patient specific approaches to EPO dosing in ESRD populations. Data developed in our labs suggests that the abundances of specific serum proteins and protein fragments correlate with EPO responsiveness in ESRD patients receiving dialysis therapy. These data suggest these proteins and protein fragments classify EPO response with more sensitivity and specificity than C-reactive protein and hepcidin. We hypothesize that specific serum proteins and peptides are candidate surrogate biomarkers for EPO responsiveness and can be used to improve existing algorithms for predictive EPO response indexing. We propose 2 aims. Aim 1 - We propose to validate selected candidate surrogate biomarkers in a larger population of ESRD patients enrolled from three geographically distinct centers. Within this aim we will address the null hypothesis that EPO resistance is characterized by altered expression of Th1- and Th2-cytokines, and address the hypothesis that serum peptides and proteins can predict EPO responsiveness. In Aim 2 - We propose to investigate the contribution of our surrogate biomarkers to dosage prediction in long-term administration of EPO and compare models base on these new biomarkers to established models. Using a novel Model Predictive Control (MPC) tool that we have previously published and demonstrated to be superior to standard EPO dosing techniques, we will access and incorporate the validated surrogate biomarkers to develop a refined clinical tool for dosing of ESAs.